Why Enterprise Architecture Management May Be Better Positioned as a Digital Twin for the Enterprises #3
A blog series by Brian Halkjær, Partner at Konfident, and Thomas Teglund, Lead Enterprise Architect at Securitas.
In the two latest installments of this blog series, we first introduced the analogy of Enterprise Architecture Management (EAM) as a digital twin for the enterprise, and secondly, we explored how modern data-driven EAM applications align with this digital twin concept.
Key Prerequisites for Data-Driven EA Success
"You can have the most sophisticated applications in the world, but if your data is wrong, your decisions will be wrong" – Thomas Teglund, Lead Enterprise Architect, Securitas.
To realize the full potential of a modern EAM application as a digital twin, it’s essential to have accurate, reliable data that is fit for purpose. Without good data, even the most advanced data-driven EAM applications cannot deliver meaningful insights. Thus, it requires both clear quality definitions and effective enforcement mechanisms to maintain a high quality of data.
Defining Good Data
Good data is accurate, complete, consistent, timely, and relevant. For EAM applications, like LeanIX, to be effective, data needs to be regularly maintained and validated. Incomplete or outdated data can lead to poor decision-making and inaccurate visualizations of the enterprise. Organizations need to establish precise standards for data modeling, naming conventions, and quality benchmarks. This ensures that everyone in the organization understands what constitutes good data and how it should be structured. In other words, we need good data governance practices.
Another key aspect in this regard is not to reinvent the same truth over and over. Look for opportunities to source as much data from existing sources as possible – and be open to discussing which new requirements this may place on data at the source, given the extended purpose of said data.
On the other hand, don’t create data silos. Ensure that you also share the information at hand, establishing an ecosystem of knowledge with trusted data. You must be particular in defining which system is the System-of-Record (SoR) for which data element, and how this is shared. This is a bigger topic, which we will explore in more detail in the next post.
Effective Enforcement (Adoption and Measurements)
Once quality standards are defined, organizations need mechanisms to consistently enforce them. This involves setting up rules, processes, and delegating responsibilities to ensure that data is continuously reviewed and aligned with the defined standards. To ensure that data remains trustworthy, complete and usable, it is critical to make data quality measurable.
Automation and Visualization of Data Quality Processes
Once we have clear definitions in place that make data quality measurable, we have paved the way for 2 key elements, which are crucial in maintaining data quality at scale:
- Automation
Automation will help us make the measurements – thereby identifying quality issues (potentially even auto correcting the issues to some extent). By automating data checks, organizations can reduce manual workload and minimize the risk of human error.
- Visualization
Once measurements are in place, visualization serves multiple purposes. The visualization can help make it actionable to improve the data quality, and it guides practitioners in improving their data maintenance habits. Furthermore, it communicates clearly to decision makers how fit for purpose the data is – hence, how trustworthy the insights are. Finally, it enables you to set goals and track progress towards these, which improves both collaboration and communication.
Jointly, this greatly improves the feasibility of maintaining a trustworthy and fit for purpose representation of the enterprise. A prerequisite for enabling strategic decisions – as they will only happen when they are known to be based on reliable, accurate data.
Up Next in the Blog Series
As hinted above, the next post will delve deeper into how a Digital Twin of the Enterprise fits into a bigger ecosystem of applications.
